Tobacco Documents Online

Papers middle-aged men and women in Emt~n Finland. ~ 1996,'93:1372-9. 20 Manaon JE, Willett WC, Stampfe~- MJ, Colditz GA, Hunte~ DJ. ~ SIL Body weight and rnor~ty among women. N ~ J M~d 1995"333:677-85. 21 Deparmaent of Health. TIt~Itmltkoft~no.tion:astrat~yfor~lO, in~ ~ London: HMSO, 1992. 92 Department of Health. Ob~t):Ra~r~/nf t/~/na-axv/ng p~b/t~ of0~)/~ 23 GarrowJ.O0~-/ty a~dn/~dd/mut London:Churchill Livingstone, 1988. 24 Bennett N, Dodd T, FlatleyJ, Freetha S, Boiling K. Hm/t~ ~ of Fa~n~ 1993. London: HMSO, 1993. 25 American Imtimte of Nutrifior~ Report of the AIN steering commiuee on healthy weight.JNutr 1994;124:2240-X (Aatp~111 ~ 1~) Prospective study of Helicobacter pylori seropositivity and cardiovascular diseases in a general elderly population Timo E Swandberg, Reijo S T'tlvis, Matfi Vuoristo, Magnus Lindroos, Timo U Kosunen Case-control and c~oss sectional studies have sug- gested that chronic infection with Helicobacterpylori is a risk factor for cardiovascular disease) ~ No prospective studies have examined this association in elderly people--those usually with the highest prevalence of Hpy/or/infection? Subjects, methods, and results We performed a prospective study on people from the Helsinki aging study.From the Helsinki cereus register random cohorts of people alive in July 1988 and born in 1904, 1909, and 1914 (300 in each group, 11.2% of the total population of 8035) were invited to participate; 795 people were alive and stin living in the do/of Hels/nki. A/together 144 refused to partidpa~ leaving 651 (81.9%) people to be examined clinically up to April 1990. Hpy/m'/IgG, IgA, and IgM antibodies were tinted by enzyme immunoassay~ in 624 subjects (repeatibility of the te~t exprexsed as inwadass correla- tion coefficient ~ = 0.95). Baseline evaluation included a postal questionnaire, a su'uanred interview conducted by public health nurses, a review of all available patient records, a clinical examination carried out by general practitioners, and laboratory investigations after an overnight fast. Standard 12 lead re.sting electrocardiographic data were classified according to the Minnesota code criteria in duplicate. The results of echocardiography have been published elsewhere? The presence of dise~e was based either on data from hospital records or on clinical examination, with special emphasis placed on the diagnostic accuracy for cardiac disease. Subjects were assigned to a group of "healthy elderly" (n--- 122) if their subjective and objec- tive (according to the examining physicians) health was good or mode~te; they did not have hyper~nsion, diabetes, dementia, or symptoms of cardiovascular, cerebrovascular, or pulmonary diseases, cancer, or other disabling diseases; and their history showed nor- mal exercise tolerance. During the five year follow up 250 subjects died. The principal cause of death was determined from the death certilkates by trained nosologists at the Central Statistical Ol~ce of Fmlan~L Data were analysed with BMDP software. The ditfer- ences in laborato~/variables were tested with an analy- sis of variance in which age and sex were included as covariate~ The Cox proportional hazards model served to test the influence of Hpy/or/seroposidvio/on survival, with age and sex used as covariate~ Among the healthy elderly, 68% were seropositive for Hpy/or/. Seropositivio/did not differ significantly in subjects with and without manifest vascular diseases (67% v 71%; difference -4.3% 95% confidence interval - 11.8% to 3.3%). The prevalence of major electrocardiographic or echocardiographic abnonnalio/in subjects seroposi- tive and seronegadve for H py/ori was similar. Of the laboratory variables, only the serum concentration of high density lipoprotein cholesterol differed between the 419 seropositive subjects and the 190 seronegative subjects (1.46 mmoiA v 155 mmolA; difference - 0.09mmol/l, -0.I7 to - 0.01; P= 0.04). During the five year follow up, crude mortality was 40~ and cardioramflar mortality 20% in the whole ser~ After age and sex were controlled for, H py/or/ seropositivio/ was not related either to all cause or cardiovascular mortality (table 1). As expected, several baseline cardknraso.dar variables (clinical symptoms and signs, electrocardiographic and echocardiogral~hic abnom~ities) significantly predicted mortality~for exam~ total mortality among subjeos with and s~¢p 1318 Division of Dep~tme~t of Medidn~ Unlve~sity of HelsinkL FIN-00~90 Helsinki, Timo E Swandberg, Reijo S ~dv~ profe~or of geria~ Ma~ Vuorlsto, Deparunent of Bacteriology and Immunology, University of H~ink~ Tano U Ir~sunen, l~-par~eat of Cardiology, Vaa~a Central Hospital. Vaasa, F'mland Magnus Lindr~. Correspondence to: Profemor T'dvis. BMJ 1997;314:1~ 17-8 Table 1 Retat~e risk (95% confidence interval) of death within five years from cardi~,ascu~ar causes a~d all causes in people aged 75-85 ~rdlmm~ular moflal~ Total mort=l~ Groul= (r~127) (n=~50) Age (5 yea~ group/rigs) L77 (1.41 to 2.22) 1.61 (1.38 to 1.88) Ma~e sex 1.16 (0.77 to 1.75) 1.59 (120 to 2.08) H py~# sero~x~,~i'~ty° 1,07 (0.73 to 1.55) 1.08 (0.83 to 1.42) * Relative risk tested with ~e Cox re~ression arm~ysis, controlled for ago and sex. 0 0

Papers without major electrocardiographic abnormalities at baseline (47% v 35%; difference 12.7%, 5.1% to 20.7%). Comment On cross sectional analysis we found no assodation between Hpy/or/seroposidvity and cardiovascular dis- eases (assessed in severai ways) in our 624 elderly sub- jects. Because H py/ori seropositivity did not predict total or cardiovascular mortality during a five year fol- low up, our results offer no evidence for an association between Hpylori infection and coronary heart disease, and they differ from those reported recendy in younger subjects)* Our results do not exclude the possibility that chronic Hpylori infection acquired early in Life may increase the lifelong risk of coronary heart disease. By analogy with serum cholesterol concentra- don in elderly subjects, conn'olled intervention studies may be needed to ascertain whether eradicating Hprior/infection in certain subgroups is worthwhile. Funding': Ragnar Ekberg Foundation and Yrj6 Jahnsson Foundation, Helsinld, Finland. Conflict of interest: Non~ Patti P, Mend,all MA, CanJngton D, Strachan D, Leatham E, Molineaux N, a aL Association of Hdicobacter pylori and ChhmydJa pneumoniae infecdom with coronary h~art disease and cardiovascular risk factom 2 Murray LJ, Bamford KB. O'Reilly DPJ, McCrum EE, Evans AE. Hdicobacter pylori infection: relation with ca~tiosascular risk factors, isdmemic heart disease, and social class. Br HeartJ 1995;74:497-501. $ Kosunen TU, H~kJ, Raumlin HI, Myllylll G. Age-dependent increase of Campylobact~r pylori an~,dies in blood dmmrs. &andJ 1989"24:110-4. 4 Lindroos M, Kupari M, HdkldllJ, T'dvis RS. Prew, dence of anrfic ~alve almormalities in the elderly: an ed~ocardiographi¢ study of a random population sample.JAm Cog C_,ardiN 199$;21:1220-5. S~ep 1317 Gastroenterology, L Sacco University Hospital. Via G B Grassi 74, 20157 Fabrizio Parent:e, G~ovanni Maconi, Omella Sangalerd, GabHele Bianchi Porr~ I~-par~ent of Haematolog); L Sacco Unive~ity Hospital, Milan Marina Poggio, r,~g/strar Edoardo RossL Department of Medical Statistics and Biometr); Univerfity of Milan Piergiorgio Duca, ~ ~0f~0~ Correspondence to: Professor Bianchi Porr~ BMJ 1997"314:1318-9 Helicobacter pylori infection and coagulation in healthy people Fabrizio Parente, Giovanni Maconi, Vencrina Imbesi, Omella Sangaletti, Marina Poggio, Edoardo Rossi, Piergiorgio Duca, Gabriele Bianchi Porro Helicobaaer pylori infection has recently been assod- ated with an increased risk of developing ischaemic heart disease~t 2 It has been suggested that chronic gas- tritls related to H pylori infection may increase, through inflammatory mediators, the concentration of certain coagulation factors such as fibrinogen,~ which are pre- dictors ofischaemic heart disease.~ We investigated the potential assodation between H py/or/infection and abnormalities of plasma coagulation in healthy people, with particular emphasis on the possibility of Hpylori inducing a tendency towards coagulation, thereby influencing the risk of ischaemic heart disease. Subjects, methods, and results Initially, 368 consecutive asymptomatic blood donors (unpaid volunteers) were recruited for this study. Exclusion criteria were age > 51 years, any chronic drug treatment, recent intake of drugs interfering with blood coagulation, use of oral contraceptives, previous treatment for H py/or/infection, pregnancy or breast feeding, and previous diagnosis of ischaemic heart dis- ease, peptic ulcer, or any systemic chronic illness. Dietary habits, alcohol and dgarette consumption, and sodoeconomic status were determined. A total of 300 subjects (229 men) aged 20-51 (mean 34.7) years fulfilled the inclusion criteria and were enrolled into the stud): A resting venous blood sample was taken in all subjects and was analysed for concentradom of total cholesterol, C reactive protein, plasma tibrinogen, factor VII C, and haemoglobin; erythrocyte sedimenta- tion rate; prothrombin time; pardal thromboplastin time; and platelet and leucocyte count. Prothrombin cleavage fragment (factors I and II), an index of prothrombin activation,~ was also assayed. IgG antibodies specific to /-/pytor/ were determined by using a commercial F_J.2~A kit (Helori test, Eurospital, Trieste, Italy); a cut off value of 19% was used, based on previous analysis of 200 patients (sensitivity compared with histology, 92%; specificity, 94%)• Student's t test and the X~ test were used to compare characteristics of subjects and values of haemostatic factors in subjects with and without Hpy/or/infection; multiple regression was used to assess the effects ofcovadates. The overall prevalence of H py/ori infection was 53% (158/300).Table 1 shows that subjects positive for Hpy/or/were significantly older than those negative for Hpy/or/. The groups did not differ significantly in other characteristics or in values for plasma fibrinogen, cholesterol, leucocyte and platelet count, myttmacyte sedimentation rate, prothrombin time, partial throm- boplastin dine, and C reactive protein. However, concentrations of factor VII C and prothrombin cleav- age fragment were significandy higher in positive than in negative subjects, though the association disap- peared after adjustment by multiple logisdc regression for age, sex, and social class. ~omment As plasma fibrinogen, and total leucocyte count, which are well known risk factors for ischaemic heart disease,~ are increased in padents infected with H py/or/) the increased risk of ischaemic heart disease in people posidve for H py/or/may be mediated through raised plasma fibrinogen concentrations. However, a large cross sectional population survey failed to find a significant association between H py/ori and fibrinogen.~ These studies may be biased because they included padents with ischaemic heart disease, a condition which could be associated with increased concentrations of coagulation factors irrespective of patients' Hpy/trri status. Comparing the concentrations of circulating coagulation factors in healthy people